Testing a Tiered Rewards Program Before Changing It

Prepared by Monroe Analytics | A real project from a previous analyst role, described by method only. Company details and figures are left out.

The situation

In a previous analyst role, I led and built the analysis of a tiered program that a company used to reward its customers for repeat buying. The more a customer bought, the higher the tier they reached, and each tier paid a bigger reward. The tier thresholds had been set once and were up for review. The people who owned the program wanted to know what would happen to participation and cost if they moved the lines.

The question

"If we make each tier easier or harder to reach, how many customers land in each tier, and what does the program cost us?"

What I did

• Started from last year's actual behavior. I used each customer's real purchase count for the year, so every scenario was tested against what customers actually did, not against a guess.
• Turned each proposal into a scenario. A scenario is just a set of thresholds, one per tier. I built a small tool that takes any set of thresholds, sorts every customer into a tier and checks the thresholds make sense (no tier easier than the one below it).
• Priced every scenario. Each tier pays a set reward, and higher tiers include the rewards below them. Multiplying customers per tier by the reward gives the total cost and the average cost per customer, side by side for every scenario.
• Compared the shape, not just the total. Two scenarios can cost the same and still push very different numbers of customers into the top tiers. I compared how customers spread across tiers so the trade-off between cost and reach was visible.
• Looked at who sits in each tier. I joined the customer tiers to the purchase-level records to see what kind of buying each tier was made of. This shows whether a tier is being earned by the behavior you actually want to reward.
• Made it repeatable. New thresholds could be run in minutes, so the owners could ask "what if" as many times as they liked in the same meeting.

What it gave the business

A clear, priced view of each proposed change before it was announced, so the decision was based on cost and reach together instead of on a single total.

Why it matters for a small business

Loyalty cards, membership levels and volume discounts all work this way. Before you move a line, you can test it on last year's customers and see the cost and the winners.

Interested in something like this for your business? Book a free 30-minute call: https://calendly.com/titus-christofferson-monroe-analytics/30min